Amtrak Crash Medical Injury Factual Report

8/19/2019 Amtrak Crash Medical Injury Factual Report

1/15

1

NATIONAL TRANSPORTATION SAFETY BOARD

Office of Research and EngineeringWashington, DC

Injury Group Chairman Factual Report

January 20, 2016

Mary Pat McKay, MD, MPHChief Medical Officer

A. ACCIDENT: DCA15FMR010Accident Type: Train DerailmentLocation: Philadelphia, PA

Date: May 12, 2015

Time: 9:21 pm (EDT)

Train#1: 188

Carrier #1: Amtrak

B. GROUP IDENTIFICATION:

Mary Pat McKay, MD, MPHGroup Chairman

Chief Medical Officer

National Transportation Safety Board

Dana Sanzo

Rail Survival Factors Investigator


Nicholas Webster, MD, MPH

Medical Officer


Kristin Poland, PhDSenior Biomechanical Engineer


Tom Barth, PhD

Highway Survival Factors Investigator


Pete Wentz

Aviation Survival Factors Investigator


C. DETAILS OF INVESTIGATION

1. PurposeThis report was developed to describe the injuries sustained by the occupants of the

Philadelphia, PA Amtrak train derailment.


2/15

Injury Group Chairman Report – DCA15MR010 2

2. Background and MethodsThe autopsy reports, accident photographs, EMS records from occupants transported

from the scene, and medical records from the injured occupants treated at area hospitals

were reviewed and combined with some information obtained from occupant interviews.

Identification of Injured Occupants

In the days following the accident, the NTSB’s Transportation Disaster AssistanceDivision compiled a list of passengers and train crew who visited local hospitals with

injuries. The information came from many sources including (but not limited to): Amtrak,

the Philadelphia Office of Emergency Management, the American Red Cross, the

Philadelphia Medical Examiner’s Office, the Philadelphia Police Department, andvarious hospitals (Temple University Hospital, Jefferson, Hahnemann, Aria Torresdale,

Einstein Medical Center, and Penn Presbyterian Medical Center).

The compiled information regarding occupants who received local hospital treatment

immediately following the accident was used to subpoena medical records as described below. No information was collected on any occupants who may have sought medicalcare outside the local area, at non-hospital locations (like urgent care centers or physician

offices), or who may have sought care after the 13th

of May.

Medical RecordsFor each identified occupant, the NTSB subpoenaed the following items from the

hospitals where treatment occurred:

(1) Complete emergency department records (physician and nursing notes,

medication administration record, procedural notes, radiology readings, laboratory

results);

(2) If admitted, the admission note, any physician transfer notes, final radiology

reports, any operation notes, and the discharge summary; and

(3) For all patients, the discharge or transfer instructions and billing records (to

include coded information such as ICD-9, E&M, or CPT codes, but may exclude

financial information).

Hospitals provided a variety of typewritten, computerized, and handwritten documents in

response to the NTSB subpoenas. In this accident, 20 subpoenaed medical records were

not obtained, generally because of discrepancies in names or dates that could not beresolved. The information for these injured persons is not included in this report.

Demographic Data and Notes

The demographic data (age, sex, height, and weight) came from information in the

medical records. In some cases, the height and weight were noted to be estimates. Train

car locations were established by interviews with the occupants by police or investigators


3/15


related to the NTSB. Notes included in Attachment A are descriptions of what happened

to the individual that were recorded in the medical records.

Injury Description and Coding

The Abbreviated Injury Scale (AIS) was used to code the injury information. 1

The AIS

system provides a scientific coding method for identifying and describing injuries.Coders require specific training in using this method. NTSB contracted with three

trained, certified AIS coders from the Army Research Laboratory (Aberdeen, MD) to

provide expert assistance and ensure valid coding using the 2005/Update 2008 version of

AIS. In addition, the AIS assigns a specific, six digit code as well as an AIS severityscore of 1 (minimal) to 6 (fatal) to each injury. When there is insufficient information to

determine the degree of injury, an AIS severity score of 9 was assigned.

For each of the injured occupants, AIS codes were assigned to each injury diagnosed by

treating health care providers, as recorded in the medical records. In some cases,

occupants complained of pain in areas where no injury was formally diagnosed. These

complaints were recorded in narrative format but no AIS code was assigned. Where therewere discrepancies, such as multiple providers describing the same injury differently, the

final radiology reading (if available) was used to determine presence or absence of

injuries. In cases where the radiology reading used words like “possible” or “clinicalcorrelation required,” notes from the physician staff and their diagnoses were used to

define whether or not the specific injury was present. Of note, no radiological images

were obtained and no direct discussion with health care providers regarding specificinjuries was carried out. In addition, AIS coding was only performed for occupants who

survived long enough to have a complete hospital evaluation of their injuries. AIS

coding was not performed for deceased occupants as their injury descriptions were lesscomplete and specific than the information obtained on survivors.

Along with each AIS code and AIS severity score, a narrative description of each injuryis provided in Attachment A: Occupant Injuries, at the end of this report. This

information includes any available specifics, such as left or right, upper or lower arm, and

the loss of height quoted on the radiology reports of spinal fractures. It also includes the

narrative description of signs or symptoms that could not be coded. Figures are includedthroughout this report that were created by the Army Research Laboratory personnel

using VisualAID, a tool developed by them to support AIS coding and results

demonstration.2

NTSB Injury Severity

The NTSB also codes people with injuries into four severity codes: fatal, serious, minor,and none. According to Title 49, Code of Federal Regulations, Section 830.2, serious

injuries are defined as an injury which is sustained by a person in an accident and which:

Requires hospitalization for more than 48 hours, commencing within

seven days from the date the injury was received;Results in a fracture of any bone (except fractures of fingers, toes, or

nose);


4/15


Involves lacerations which cause severe hemorrhage, nerve, muscle, or

tendon damage;

Involves injury to any internal organ;Involves second or third-degree burns, or any burns affecting more than

five percent of the body surface;

If a person dies within 30 days as a result of injuries sustained in an NTSB investigated

accident, they are coded as “fatal.” Survivors with injuries that do not meet the “serious”

definition are coded as “minor.” If no injury diagnosis is made, the person is coded as

having no injury.

Non-injury diagnoses such as pre-existing co-morbid diseases (like high blood pressure

or diabetes), complications of hospitalization (such as pneumonia, blood clots, or urinarytract infections), and psychological trauma were not included.

Injury Severity Scores

In addition, an Injury Severity Score (ISS) was calculated for each occupant whoseinjuries were AIS coded. The ISS predicts the likelihood of survival among traumatically

injured patients and can be used to compare the severity of injury among individuals. In

order to calculate the ISS, the AIS coded injuries are divided into six regions: 1)head/neck, 2) face, 3) chest, 4) abdominal and pelvic contents, 5) extremities and pelvic

bones, and 6) external. ISS is then calculated by adding together the squares of the

highest AIS scores for each of the three highest scoring body regions:

ISS = (highest AIS regionA)2 + (highest AIS regionB)

2 + (highest AIS regionC)

2

The maximum survivable ISS is 75.3 Injuries with unknown severity are coded to a

severity of 9. If there were other known injuries, the injury coded to 9 was removed and

an ISS calculated using known severities for the other injuries.

Injury severity scores are routinely divided into four groups: minor (ISS 1-8), moderate

(ISS 9-15), sever e (ISS 16-24), and very severe (ISS >25); the results below are presented

in these groups.4

Emergency Medical Response and Victim Transportation

The Philadelphia Fire Department (PFD) provided the NTSB with a copy of theoperational procedure for mass casualty incidents in place at the time of the accident

(OPS#35-11, See Attachment B; Attachment C is the updated version of OPS#35

developed following this accident). According to this document, an incident of this size(requiring at least 30 ambulances) meets the PFD classification for a mass casualty

disaster.

According to the operational procedure, injured persons in a large mass casualty incidentwere to be triaged into four categories: Priority 1 (RED), Priority 2 (YELLOW), Priority

3 (Green) and DEAD (BLACK). A color coded triage tag was to be affixed to each

injured person to identify that they had been triaged and their triage category. An


5/15


individual identified as the Patient Transportation Group Supervisor was to, “Ensure that

victims are transported to appropriate hospitals. Whenever possible, Priority 1 (Red) Tag

patients should be distributed to trauma centers. If there are entrapped Priority 1 (Red)tag patients, consideration may be given to holding beds at the closest trauma center(s)

for them. Priority 3 (Green) tag patients can be transported to more distant hospitals. Do

not overload any individual hospital, if possible. (Section 3.18.1g)”

In addition, an individual was to have been identified as the Medical Communications

Coordinator. This person’s responsibility was to establish the status of hospital/medical

center availability and communicate that information with the Transport Coordinator inorder to ensure no one institution was overwhelmed and that patients were distributed

appropriately among available resources.

The PFD operational procedure describes using available PFD ambulances for transport

of the injured, followed by non-municipal ambulances with an existing memorandum of

understanding, followed by out-of-county ambulances, with a procedure for vetting each

type of transport. The procedure allows for use of non-ambulance mass transit vehiclesincluding Septa buses, if needed (PFD OPS#35 2011, Addendum #2 in Attachment B).

Trauma Center DesignationIn Pennsylvania, the Pennsylvania Trauma Systems Foundation

5 accredits hospitals as

trauma centers, separated into four distinct levels (I-IV). Level I Trauma Centers provide

multidisciplinary treatment and specialized resources for at least 600 major trauma patients per year and must perform trauma research and train surgeons. Level II trauma

centers provide similar medical services to at least 350 major trauma patients per year but

do not perform the research and training components. Level III trauma centers may carefor moderately injured trauma patients and both Level III and IV trauma centers have the

capacity to stabilize and transfer seriously injured patients to a higher level of trauma

care. Some hospitals may be able to provide trauma care but have not undergone theformal process to obtain a trauma designation.

Transport of Injured Patients by Police in Philadelphia

In Philadelphia, since 1987, a policy has been in place that allows police department personnel to transport victims of penetrating trauma to trauma centers. In part, this

practice has been supported by research performed in Philadelphia that demonstrated at

least equivalent survival outcomes for victims of penetrating trauma transported by policeand EMS between 1986 and 1992

6 and again from 2003 to 2007.

7 According to this

literature, published by physicians practicing in Philadelphia, there are hundreds of police

transports of injured patients to emergency departments each year, with little to nomedical care provided en route.7


6/15


3. Results

Occupant Injury Severity

Medical treatment records or autopsy reports were obtained and reviewed for a total of

172 train occupants. Of these, eight passengers died as a result of the accident, 43 met the

NTSB definition of serious injury. An additional 113 occupants had minor injuries. Atotal of 7 occupants were evaluated without any diagnosis of an injury, and one left the

hospital without being seen and is included as not having an injury. (See Table 1.) These

occupants ranged in age from 12 to 79 with a median age of 39 year (quartiles 30, 54)and 88 of them (51.2 percent) were male.

Table 1. Occupants NTSB Injury Severity

Number Percent

Fatal 8 4.6

Serious 43 25.0

Minor 113 65.7

None 8 4.7

Total 172 100.0

No AIS coding was performed for the eight fatally injured passengers or the eight

occupants who were uninjured. Eight other occupants’ medical records lacked sufficientinformation to calculate an injury severity code (ISS), either because no diagnosis was

made or because the AIS code(s) assigned had a severity level of 9 (unknown). (See

Table 2.) The ISS score ranged from 0 to 57.

Table 2. Injury Severity Scores for Train Occupants

ISS Severity Number PercentMinor (1-8) 124 72.0

Moderate (9-15) 12 7.0

Severe (16-24) 6 3.5

Very Severe (>25) 6 3.5

ISS not calculated 24 14.0

Total 172 100

Injuries by Train Car LocationTrain car locations for 80 occupants were obtained by interview. Some occupants were

unable to give an interview or unwilling to speak with investigators. Others left the area before providing information. Some occupants did not know their location and otherswere unsure of their exact location on the train. If they were unsure, this uncertainty is

reflected in Table 3, below. Eleven people held business class tickets and were presumed

to have been riding in the first passenger car, which was the only business class car in thetrain consist.


7/15


Table 3. Occupant NTSB Injury Severity by Train car (if known)

NTSB Injury Severity

Train Car Fatal Serious Minor None Total

Locomotive 0 0 1 0 1

1 3 7 1 0 11

2 0 5 6 1 12

2 or 3 0 0 1 0 1

3 0 3 12 0 15

3 or 4 0 0 2 0 2

4 0 1 6 1 8

4 or 5 0 0 1 0 1

5 0 2 8 0 10

5 or 6 0 0 2 0 2

6 0 1 7 0 8

6 or 7 0 0 3 0 3

7 0 3 14 0 17

Unknown 5 21 49 6 81

Total 8 43 113 8 172

Emergency Medical Response

Seven train occupants did not survive and were not transported from the scene. One person was transported to Temple Hospital and died in the emergency department. The

large majority of the injured were transported to the hospital by the police department

using police cars, police vans, and paddy wagons. Septa buses were also used to transportsome of the injured. Some train occupants may have been triaged by emergency medicalservices (EMS) personnel before being transported by police or bus but none had hospital

emergency treatment record that reflected an EMS mass casualty triage tag was present

on arrival.

The Philadelphia Fire Department provided NTSB with ambulance (EMS) transport

charts for 12 train occupants transported from the scene to a hospital emergencydepartment by ambulance; these included records of one person who died at the hospital.

In addition, they reported another 12 people were transported by EMS but the charts for

these transports were missing. Without a record of transport including the occupant’s

name, these could not be combined with the injury information from the hospital records.


8/15


Table 4 shows the relationship between having an EMS transport chart (for those 12

cases where EMS charts were available) and the NTSB definition of injury severity whileTable 5 demonstrates the relationship with ISS (among those with a calculated ISS).

Table 4. Transportation to Hospital by NTSB Injury SeverityFatal Serious Minor None Total

EMS Transport Chart 1 3 7 1 12

No EMS Chart 0 40 106 7 153

Total 1 43 113 8 165

Table 5. Transportation to Hospital by Injury Severity Score (ISS)

Minor

(1-8)

Moderate

(9-15)

Severe

(16-24)

Very

Severe

(25+)

Total

EMS

TransportChart

7 1 0 2 10

No EMS Chart 117 11 6 4 138

Total 124 12 6 6 148

Train occupants were transported to ten local hospitals from the accident scene; those

hospitals are identified by their trauma level in Table 6.

Table 6. Hospitals’ Trauma Center Level Designation

Hospital Name Trauma Center

Designation

Level

Aria Frankford none

Aria Torresdale II

Einstein I

Episcopal none

Hahnemann I

Holy Redeemer none

Hospital of the

University of

Pennsylvania (HUP)

none

Jefferson I

Penn Presbyterian* I

Temple I*According to the Pennsylvania Trauma Systems

Foundation, the Level I trauma center designation moved

within the Penn medical system at HUP to Penn

Presbyterian on February 4, 2015.


9/15


Occupants were identified by the hospital who provided records. In three cases, those

records described a transfer from an outside (previous) hospital but it was not clear whichhospital had transferred the injured occupant. Tables 7 (NTSB severity) and 8 (ISS)

demonstrate the injury severity according to the hospital primarily treating the occupant.

Figure 1 demonstrates the ISS information graphically.

Table 7. Treating Hospitals by NTSB Injury Severity

Trauma

Center

Designation

Level

Fatal Serious Minor None Total

Aria Frankford none 0 1 14 0 15

Aria Torresdale II 0 4 18 3 25

Einstein I 0 2 5 0 7

Episcopal none 0 0 10 0 10

Hahnemann I 0 11 17 3 31

Holy Redeemer none 0 0 5 0 5

HUP none 0 0 3 0 3

Jefferson I 0 8 14 2 24

Penn

Presbyterian

I 0 2* 0 0 2

Temple I 1 15* 27 0 43

Total 1 43 113 8 165* Two people at Penn Presbyterian and one person treated at Temple were initially seen at

another emergency department and transferred.


10/15


Table 8. Treating Hospitals by Injury Severity Score

Trauma

Center

Designation

Level

Minor

(1-8)

Moderate

(9-15)

Severe

(16-

24)

Very

Severe

(25+)

Total

Aria

Frankford

none 14 0 0 0 15

Aria

Torresdale

II 18 1 0 2 21

Einstein I 7 0 0 0 7

Episcopal none 10 0 0 0 10

Hahnemann I 21 3 2 1 27

Holy

Redeemer

none 5 0 0 0 5

HUPnone 3 0 0 0 3

Jefferson I 17 1 1 0 19

Penn

Presbyterian

I 1* 0 0 1* 2

Temple I 28 6 3 2* 39

Total 124 12 6 6 148* Two people at Penn Presbyterian and one person treated at Temple were initially seen at

another emergency department and transferred.

Figure 1. Number of Injured and their Injury Severity Scores at the Treating Hospitals

Trauma Centers


11/15


Time of Arrival to Emergency Departments

The accident occurred at 9:21 pm local time. Specific transport times could not bedetermined for the majority of transported train occupants because they did not have

EMS charts. As a result, the initial time stamp present in occupants’ hospitals records was

used to describe the time of arrival.a Three occupants were transferred, and, as a result,the initial time stamp at the first hospital was used for this description.

For 6 surviving occupants, the initial time of arrival could not be determined from

available records, and for two transferred occupants, the initial hospital location could not be determined. These individuals are not included in the following description. Among

the remaining 156 occupants, the first time stamp was at 9:57 pm (36 minutes after the

accident) and the last person arrived at a hospital at 1:02 am on May 13th

, 2015 (3 hours,41 minutes after the accident). The median time of hospital arrival was 11:15 pm (25

th

quartile 10:47 pm and 75th

quartile 11:35 pm). Table 9 demonstrates the timing of arrivals

at each hospital by hour. Overall, among the injured occupants transported to hospitals

with available data, 139 (89.1%) had been registered into an emergency department bymidnight, about two and a half hours after the train derailed.

Table 9. Time of Hospital Arrival, by Hour

Before

11 pm

11pm -

midnight

After

midnight

Total

Aria Frankford 13 2 0 15

Aria Torresdale (TC) 2 18 4 24

Einstein (TC) 5 2 0 7

Episcopal 5 1 4 10Hahnemann (TC) 6 21 3 30

Holy Redeemer 0 2 3 5

HUP 0 2 1 3

Jefferson (TC) 5 17 0 22

Penn Presbyterian (TC)* 0 0 0 0

Temple (TC) 23 15 2 40

Total 59 80 17 156

* Both occupants treated at Penn Presbyterian were transferred from an initial hospital.

Figure 2 is a map based description of the number and severity of the victims at each

hospital at approximately 1:30 am on May 13th

.

a Because of the need to register patients into the hospital, there may be a delay between patients’ physical arrival

time and the initial time stamp in the medical record; this may be longer if multiple patients arrive simultaneously.


12/15


Figure 2. Map of Relative Numbers of Patients Treated in Hospitals at 1:30 am

InjuriesThe 156 injured but surviving occupants with medical records available to the NTSB had

a total of 597 individually described injuries; of these, 538 were AIS coded. A description

of each injury including those to the deceased occupants is provided in Attachment A tothis report.

Table 10, below, outlines the AIS coded injuries by their ISS body region. In the headand neck region, there were 23 head injuries where the severity could not be coded. All of

these were cases where the healthcare provider had simply noted “head injury” without

further descriptors or diagnosis. In each of these cases, the occupant had undergone a CTscan that was negative for any intracranial pathology. The remainder of the head and

neck injuries included cervical strains, concussions, head injuries with headaches, and

one linear skull fracture (AIS 3). In the cervical spine, there were six transverse processfractures as well as one articular process and one facet fracture. One person had bilateral

jumped cervical facets and a spinal cord injury (AIS 5). No survivors had any significantintracranial bleeding or swelling.

Trauma CenterNon-trauma Hospital


13/15


Table 10. Injuries by Body Region

The occupants’ facial injuries consisted primarily of fractured noses and teeth. There wasone corneal abrasion and one fracture of the posterior wall of the maxillary sinus.

Twenty four survivors had 68 AIS level 2 or higher chest injuries including people withflail chests (3), pulmonary contusions (9), multiple rib fractures (15), and a fractured

sternum (1). There were hemothoraces (3), pneumothoraces (11), and two

hemopneumothoraces. In addition, there were two lung lacerations and one ruptured

diaphragm. There were 15 fractures of the spinous process or transverse process of thethoracic vertebrae (in 6 people) and 6 fractures of the endplate of the thoracic vertebrae

(in four people) but no vascular injuries in the chest and no thoracic spinal cord injuries.In addition, one person was diagnosed with contused ribs and one with a contusion of thesternum.

Among the injuries to abdominopelvic contents, there were 22 lumbar spine fractures but

these included 19 transverse process fractures, one spinous process fracture, one fractureof the anterior superior corner, and a single 30% compression fracture of the body of the

vertebra. In addition, there was one extraperitoneal bladder rupture, one liver contusion,

three splenic lacerations, two kidney injuries, one retroperitoneal hemorrhage from anunknown source, three bowel injuries, and a lumbar strain.

The extremity and pelvic injuries were primarily fractures and dislocations, some ofwhich were open. The two AIS level 5 injuries were two very severe pelvic fractures.

The five AIS level 3 injuries included a pelvic fracture, two complex ankle fractures (one

open), and two femur fractures.

Overall, 43 victims fractured at least one bone. Figure 3 demonstrates the areas offracture (note, left and right are not accounted for in this figure).

Body Region AIS Severity

1 2 3 4 5 9 Total

Head/neck 20 12 1 0 1 23 57

Face 7 2 0 0 0 0 9Chest 2 35 26 6 1 0 70

Abdominal/pelvic contents 3 32 2 1 0 0 38

Extremities/pelvic girdle 31 36 5 0 2 0 74

External 284 6 0 0 0 0 290

Total 347 123 34 7 4 23 538


14/15


Figure 3. Heat Chart of Fractures by location

The external injuries were primarily abrasions and contusions; those that were AIS level

2 severity were lacerations that were very large or had arterial bleeding described.

Overall, 48 train occupants had injuries at AIS level 2 or higher (Figure 4). These

diagrams highlight that the majority of serious injuries in the survivors in this accident

were to the chest; there were few in the head and neck.

D. SUMMARY OF INJURY FINDINGS With 253 people on the train at the time of the derailment, 7 died of injuries at the scene, one

died in the emergency department, and 185 were reported to have been seen in 10 local hospitals.

Of these, the NTSB obtained medical records for 165 surviving train occupants and reviewed theautopsy findings of the eight deceased occupants. Overall, among the survivors in this accident

there was one moderately severe head injury (AIS 3) but no intracranial hemorrhages, a single

cervical spinal cord injury (AIS 5), and twenty four occupants with 68 significant (AIS 2+) chestinjuries.

Review of the records demonstrated the large majority of injured occupants were transported for

medical care by police vehicle or Septa bus. Only 3 of 43 people with serious injuries had anambulance transport chart. The injured were distributed unevenly across nearby Trauma Centers

and local hospitals.


15/15


E. ATTACHMENTSAttachment A: Occupant Injury Descriptions, Amtrak Derailment, 2015

Attachment B: Philadelphia Fire Department Operational Procedure: Multiple Patient /

Mass Casualty Incidents (June 2011)

Attachment C: Philadelphia Fire Department Operational Procedure: Multiple Patient /

Mass Casualty Incidents (September 2015)

References

1 AIS 2005 Update 2008. Association for the Advancement of Automotive Medicine, Barrington, IL 2008.2 Visual Anatomical Injury Descriptor (VAID), Version 3.9, Army Research Laboratory,

United States Department of Defense.3 Baker SP, O.N.B., Haddon W, et al., The Injury Severity Score: a method for describing patients with multiple

injuries and evaluating emergency care. J Trauma, 1974. 14: p. 187-196.4

American College of Surgeons. National Trauma Data Bank 2013 Annual Report. Accessed 12/13/2015.Available from: http://www.facs.org/trauma/ntdb/pdf/ntdb-annual-report-2013.pdf 5 Pennsylvania Trauma Systems Foundation. What is a Trauma Center? http://www.ptsf.org/index.php/our-trauma-

centers/whats-trauma. Accessed 1/8/2016. 6 Branas CC, Sing RF, Davidson SJ. Urban trauma transport of assaulted patients using nonmedical personnel.

Acad Emerg Med. 1995;2(6):486-93.7 Band RA, Pryor JP, Gaieski DF, Dickinson ET, Cummings D, Carr BG. Injury-adjusted mortality of patients

transported by police following penetrating trauma. Acad Emerg Med. 2011;18(1):32-7.
http://www.facs.org/trauma/ntdb/pdf/ntdb-annual-report-2013.pdfhttp://www.facs.org/trauma/ntdb/pdf/ntdb-annual-report-2013.pdfhttp://www.facs.org/trauma/ntdb/pdf/ntdb-annual-report-2013.pdfhttp://www.ptsf.org/index.php/our-trauma-centers/whats-trauma.%20Accessed%201/8/2016http://www.ptsf.org/index.php/our-trauma-centers/whats-trauma.%20Accessed%201/8/2016http://www.ptsf.org/index.php/our-trauma-centers/whats-trauma.%20Accessed%201/8/2016http://www.ptsf.org/index.php/our-trauma-centers/whats-trauma.%20Accessed%201/8/2016http://www.ptsf.org/index.php/our-trauma-centers/whats-trauma.%20Accessed%201/8/2016http://www.ptsf.org/index.php/our-trauma-centers/whats-trauma.%20Accessed%201/8/2016http://www.facs.org/trauma/ntdb/pdf/ntdb-annual-report-2013.pdf

Documents

Amtrak Crash Medical Injury Factual Report